From Wiki Journal Club
Jump to: navigation, search
Kirchhof P, et al. "Early rhythm-control therapy in patients with atrial fibrillation". The New England Journal of Medicine. 2020. 383(14):1305-1316.
PubMedFull textPDFClinicalTrials.gov

Clinical Question

In patients with recently diagnosed atrial fibrillation, can early rhythm control reduce the risk of CV mortality, stroke, HF hospitalization, or ACS hospitalization?

Bottom Line

In patients with recently diagnosed atrial fibrillation, an early rhythm-control strategy was associated with a reduced risk of CV mortality, stroke, HF hospitalization, or ACS hospitalization compared to usual care. This trial did not assess withdrawal of anticoagulation or other AF medications following early rhythm control, however.

Major Points

The 2002 AFFIRM trial randomized seniors with AF to rate vs. rhythm control strategies. It found a trend towards increased mortality with a rhythm control strategy (P=0.08). Mortality was highest among subgroups that were older, had a history of CAD, or without prior HF. Whether there was a benefit from early earlier rhythm control in AF was unknown. Additionally, the rhythm control arm was allowed to stop anticoagulation if the patient remained in normal sinus rhythm for at least 4 consecutive weeks, but rates of ischemic stroke were similar between both groups. Additionally rhythm control agents were predominantly sotalol or amiodarone, which have high rates of adverse effects compared to newer agents such as dronedarone.

The Early Treatment of Atrial Fibrillation for Stroke Prevention Trial 4 (EAST-AFNET 4) randomized patients with recently diagnosed AF (defined as diagnosed ≤12 months before enrollment) to early rhythm control vs. usual care. The patients were similar in age to those in the AFFIRM trial, but started on treatment earlier. By two years 19.4% of patients in the rhythm-control group underwent ablation. This trial was stopped early with 5 years of follow-up because of efficacy in the early rhythm control arm. Specifically, the early rhythm control group had a lower risk of the CV mortality, stroke, HF hospitalization, or ACS hospitalization, as well as the individual outcomes of CV mortality and stroke. Of note, both groups continued other guideline-based AF therapies, including anticoagulation. Potential etiologies for the improvement in outcomes may be related to early AF therapy to prevent negative remodeling effects related to arrhythmia. Additionally, there were higher rates of flecainide and dronedarone use in this trial compared to AFFIRM, and lower rates of sotalol use. Overall, this trial supports the use of early rhythm control strategies for atrial fibrillation over rate control strategies. Whether anticoagulation may be safely discontinued among patients receiving early rhythm control is unclear.


As of January 2021, no guidelines have been published that reflect the results of this trial.


  • Multicenter, open, parallel-group, randomized, blinded-outcome-assessment trial
  • N=2789
    • rhythm control (n=1395)
    • rate control (n=1394)
  • Setting: 135 centers in 11 European countries
  • Enrollment: 2011-2016
  • Median follow-up: 5.1 years, stopped early at third interim analysis
  • Analysis: Intention-to-treat
  • Primary outcome: CV mortality, stroke, HF hospitalization, or ACS hospitalization


Inclusion Criteria

  • Atrial fibrillation diagnosed ≤12 months before enrollment
  • Age ≥75 years with prior TIA or stroke or ≥2 of following:
    • Age ≥65 years
    • Female sex
    • Heart failure
    • Hypertension
    • DM
    • CKD MDRD stage 3 or 4
    • LVH (diastolic septal wall width >15 mm)

Exclusion Criteria

  • Life expectancy <1 year
  • Substance use disorder
  • Prior AF procedure (ablation or surgery)
  • Previous therapy daily re on amiodarone
  • Prosthetic mitral valve or severe mitral stenosis
  • clinically relevant hepatic, thyroid, or renal dysfunction

Baseline Characteristics

From the early rhythm control group.

  • Demographics: Age 70 years, female sex 46%
  • Anthropometrics: BMI 29 kg/m2, BP 136/81 mm Hg
  • Type of atrial fibrillation: First episode 38%, paroxysmal 36%, persistent 27%
    • Median days since AF diagnosis: 36
    • Absence of AF symptoms: 30%
    • Prior cardioversion: 40%
    • CHA2DS2-VASc score: 3.4
  • Comorbidities: Prior stroke 12%, MCI or worse cognition 44%, hypertension 88%, HF 28%, CKD MDRD stage ≥3 44%

Comparing early rhythm vs. usual care at the end of the baseline visit.

  • Medications:
    • Anticoagulation: 91% vs. 90%
    • Digoxin: 3% vs. 6%
    • Beta blocker: 76% vs. 86%
    • ACE-inhibitor or ARB: 69% vs. 70%
    • Mineralocorticoid: 6% vs. 7%
    • Diuretics: 40% vs. 41%
    • Statin: 45% vs. 41%
    • Platelet inhibitor: 16% vs. 16%


  • Patients who were thought to be eligible for either rate or rhythm control were randomized to a group:
    • Early rhythm control - Treatment with an antiarrhythmic medication or AF ablation, and cardioversion of persistent AF. Medication choice for the antiarrhythmic was chosen by the local investigators using protocol guidance. Participants collected a single-lead EKG twice weekly when symptomatic. Any recorded AF triggered an in-person visits from the study team for drug escalation.
    • Usual care - Treated with rate control without rhythm control, except in the scenario of rate control for uncontrolled AF.
  • See Figure 1 on pg 1309 for specific interventions by arm at baseline and year 2, but in brief, approximately 66% of early rhythm control received rhythm control at year 2, compared with 15% of the usual care group. Listed rhythm interventions included AF ablation, dronedarone, amiodarone, flecainide, propafenone, and "other" antiarrhythmics.
  • The treatment of other CV conditions and administration of anticoagulation and rate control was continued in both arms.


Comparisons are early rhythm control vs. usual care. P-Y is person-years.

Primary Outcomes

CV mortality, stroke, HF hospitalization, or ACS hospitalization
3.9 vs. 5.0/100 P-Y (HR 0.79; 96% CI 0.66-0.94)
Stroke, death, or SAE of special interest and thought to be related to rhythm control
This was the primary safety outcome.
16.6% vs. 16.0%

Secondary Outcomes

CVD mortality
1.0 vs 1.3/100 P-Y (HR 0.72; 95% CI 0.52-0.98)
0.6 vs 0.9/100 P-Y (HR 0.65; 95% CI 0.44-0.97)
Hospitalization with worsening of heart failure
2.1 vs 2.6/100 P-Y (HR 0.81; 95% CI 0.65-1.02)
Hospitalization with ACS
0.8 vs 1.0/100 P-Y (HR 0.83; 95% CI 0.58-1.19)
Annual days admitted to the hospital
6 vs. 5 days (IRR 1.08; 95% CI 0.92-1.28)
Outcomes at 2 years
Change in LVEF: +1.5% vs. +0.8% (mean difference 0.23; 95% CI -0.46 to 0.91)
Change in EQ-5D score: -1.0 vs. -2.7 (mean difference 1.07; 95% CI -0.68 to 2.82)
Change in SF-12 mental score: +0.7 vs. +1.6 (mean difference -1.20; 95% CI -2.04 to -0.37)
Change in SF-12 physical score: +0.3 vs. +0.1 (mean difference 0.33; 95% CI -0.39 to 1.06)
Change in MoCA score: +0.1 vs. +0.1 (mean difference -0.14; 95% CI -0.39 to 0.12)
Sinus rhythm: 82% vs. 60% (OR 3.13; 95% CI 2.55 to 3.84)
No symptoms: 74% vs. 73% (OR 1.14; 95% CI 0.93 to 1.40)

Subgroup Analysis

Treatment effects for rhythm vs usual care did not vary significantly by age, sex, diabetes status, history of heart failure. There was a trend toward better outcomes in patients with CHA2DS2-VASc Score ≥4 compared to <4.

Adverse Events

Statistical comparison between groups was not provided. Please see the full list of safety outcomes in Table 3 on page 1313.

2.9 vs. 4.4%
9.9% vs. 11.8%
SAE of special interest and thought to be related to rhythm control
4.9% vs. 1.4%


  • Not blinded and didn't compare the efficacy of the given treatments.
  • This study did not withdraw anticoagulation from either arm, so it does not inform continuation or discontinuation of this high-risk therapy among individuals who have undergone early rhythm control.


  • German Ministry of Education and Research
  • Atrial Fibrillation Network (AFNET)
  • European Heart Rhythm Association
  • Leducq Foundation
  • British Heart Foundation
  • Industry support

Further Reading